A number of different systems exist for generating graphic images which may be described as three-dimensional. These systems convert data corresponding to three-dimensional images to data corresponding to two-dimensional images including depth cues such as shading, perspective, or hidden line removal. These depth cues provide three-dimensional qualities to the two-dimensional images being displayed. Very effective three-dimensional qualities can also be provided with stereoscopic displays which use binocular disparity as their primary depth cue.
In all of these systems, it is frequently desirable to provide a cursor for indicating particular positions within the images displayed. Cursors consisting of very simple constructs such as cross marks have typically been used. These constructs provide very limited depth information. Consequently, observers of the displays have often found it difficult to properly register the position of a cursor within a three-dimensional image.